Multi-UAV Task Allocation: A Team-Based Approach

This paper presents a team-search based decentralized task allocation scheme for multiple homogeneous unmanned aerial vehicles (UAVs) to provide protection to static convoys of ground vehicles. The UAVs, during operation, protect the ground convoy by searching their vicinity for imminent threat, analyzing/ confirming threat level, attacking it and finally assessing the damage to confirm if the threat has been nullified. The proposed approach utilizes search maps to form a common information base for intelligent decision making. A decentralized scheme is developed based on team theory, wherein the best course of action for each UAV is selected to minimize resource use of the team. This scheme is generic enough to handle different types of UAVs and control technique and caters to a dynamic environment. The proposed convoy protection scheme is evaluated by software simulation with multi-UAV-multi-targets. Different experiments were performed to analyze the efficacy of this approach. The performance comparison with greedy task allocation highlights the advantage of the proposed scheme.